The DAISY study and improved understanding of autoimmunity directed against insulin and novel methods to detect unknown pathogens (e.g. Skip Virgin: "sentinel immunodeficient mice and representational difference analysis for newly acquired DNA and RNA) we believe provide unique opportunities to search for infectious agents potentially triggering the development of TIDM. We propose to pursue in parallel a general search for infectious pathogens and a focused project concentrating on a "pathogenic" insulin peptide. Our preliminary data from the DAISY study indicates we can now identify children at birth (siblings of patients with type 1 diabetes with DR3/4, DQ8/2) with a risk of activating anti-islet autoantibody expression exceeding 50% within the first two years of life. In addition, we have recently discovered that given an appropriate genetic background (e.g. based on the MHC of Balb/c mice) we can readily induce anti-insulin autoantibodies with a specific insulin peptide in normal mice that is also recognized by human T cells. Using this peptide, combined with a viral mimic (poly-IC[polyinosinic-polycytodylic acid]) we can induce insulitis, and with a immunoregulatory defect in the islets (transgenic B7-1 expression), can induce islet destruction and diabetes. Development of diabetes in this model appears to be critically dependent upon genetic background, inducing environmental factors, and the "coincidental" timing of inducing factors. This provides a model for "environmental" induction of type 1 diabetes development and in the current proposal we will 1. Develop methodology and test intensive prospective evaluation of children for the appearance of anti-islet autoantibodies. 2. Detect the occurrence of infections (assays for inteferon alpha, 2',5'- oligoadenylate synthetase, "sentinel immunodeficient mice", subtractive DNA and cDNA hybridization) and sequence environmental pathogens "temporally" associated with appearance of anti-islet autoantibodies. 3. Define mimotopes of the B:9-23 peptide using "random" peptide libraries and B:9-23 responsive T cell clones. 4. Test insulin, B:9-23 mimotopes, for the induction of insulin autoantibodies, insulitis, and diabetes in genetically susceptible murine strains. 5. Evaluate peptides of infectious pathogens with B:9-23 or mimotope related sequences for ability to stimulate B:9-23 reactive T cell clones of man and mouse. 6. Test pathogen related peptides that activate B:9-23 reactive T cell clones for disease induction in genetically susceptible murine strains and evaluate as in specific aim 1, whether specific pathogens with B:9-23/mimotope related sequences infect prediabetic children and the timing of appearance in children developing anti-islet autoimmunity.